[Ashfaq]

I think you're onto something Zee. Instead of measuring a players by the runs he collect and wickets he take, we could, in a sense capture his cricketing skills through a set of diversified parameters in a background which can be tuned for all condition. I mean, if all possible condition of the gameplay could be captures in a tensor, it would be possible to detect whether the sudden flood of runs or the sudden collapse is due to change in weather. I'm thinking of a matrix that can include from soil composition of pitch taken every hour to the turbulence of wind motion measured at the batting crease every five minutes.

Given this backdrop, we could then measure the attributes of a batsman in a highly compartmentalized way. By attributes, we could say precognition, reflex, placement, economy of movement etc. One parameter should be added to measure how often a batsman can make the bowler bowl to his strength. For example, if a batsman likes to play in the midwicket (as ascertained by previous data), he must make the bowler bowl fuller. And the frequency of fuller balls in every ten balls should be compared with the average frequency of delivery. The difference should measure the influence of the batsman.

Similarly, a bowler must be measured by his consistency and variation. A weight should be added to indicated the condition of the ball. I'm out of ideas about how to correctly imprison bowlers in a matrix, though.

If this really happens, then cricket will no longer be played in stadiums. By invoking similer conditions and by simulating a player's skill matrix, a game could be decided in a simulator. Of course, to account for such huge amount of data, we'd need a supercomputer.